$NVDA

$NVDA · Tom Nash · 2026-11-13

AI 摘要 · Tom Nash

Tom Nash 长期仍极度看好 NVDA:黄仁勋的辞职是暂时的阵痛,而非逻辑破裂,因为 AI 需求、订单/积压、CUDA、网络、人才与路线图均未改变。NVDA 目前很便宜,远期市盈率低于 14,股价较 10 月高点低 35%;预计短期会有波动,出口/司法部调查是主要风险。只有当营收增长逐季放缓、利润率趋势性收缩或客户流失时,他才会改变看法。

看涨他说的是几年内
《The UNTHINKABLE is About to Happen to NVIDIA (Emergency Update)》
Tom Nash发布 2026-11-13 · 66 段内容

播放时,YouTube 可能接收设备与播放信息并使用 Cookie。 隐私说明 · Google 隐私政策

00:00/ —打开原视频

逐段跳播

66 段内容
0:0035:10

Jensen 出局了。这个真正从 Denny's 的一个卡座起步创办 Nvidia、并将其打造成一家四万亿美元公司的人,即刻离任。

Jensen is out. The man who literally started Nvidia from a Denny's booth and turned it into a four trillion dollar company is out effective immediately.

他不再是 CEO 了。他没有董事会席位。他基本上已经离开了。这就是一周之内被曝出三起据我们所知的婚外情时会发生的状况。其中一起涉及一名被指为中国间谍的人,另一起涉及一名下属。

He's no longer the CEO. He doesn't have a board seat. He's pretty much gone. That's what happens when you have three so far as we know it affairs reported in one week. One of them with an alleged Chinese spy, one of them with a subordinate.

如果 Jensen 有时间管理一家四万亿美元的公司,同时还有三段并行的关系,那我们谁都没有借口说“哦,我没时间”。如果这个人有时间同时经营三段关系和这家公司,而且显然还出色地成为世界上最优秀的 CEO 之一,那我们谁都没资格再用“哦,我就是没时间”这个借口。

if Jensen had time to manage a $4 trillion company and at the same time have three different parallel relationships, none of us have ever have the excuse of, "Oh, I don't got the time."

If this man had time to run three relationships and this company at the same time and excel apparently at being one of the best CEOs in the world, none of us ever get to use the excuse, "Oh, I just don't have the time."

周四,我们看到该股在单个交易日内一路下跌了 16%,随后跌幅收窄,最终收跌约 4.4%。

on Thursday, we saw the stock drop all the way down 16% in a single session, which mellowed down and we ended up with about 4.4% drop.

我的收件箱简直炸了。所有人都在问我同一个问题。Tom,我该怎么办?我该卖吗?我该持有吗?我该加仓吗?即便没有 Jensen,我依然极度看好 Nvidia。

My inbox is literally on fire. Everybody's asking me one question. Tom, what do I do? Do I sell? Do I hold? Do I buy more? I remain extremely bullish on Nvidia even without Jensen.

Microsoft、Google、Amazon、Meta,AI 领域世界上每一家公司都在向 AI 基础设施投入数千亿美元。它们全都是 Nvidia 的大客户。它们没有一家因为 Jensen 的离开而取消自己的计划。

Microsoft, Google, Amazon, Meta, every single company in the world in the AI sector is literally pouring hundreds of billions of dollars into the AI infrastructure. All of them are heavy clients of Nvidia.

None of them are cancelling their plans because Jensen is leaving.

对算力的需求并没有因为这家伙管不住自己的裤子而消失。科技也没有因为他有婚外情而变得过时。Kura 还在。网络效应还在。Nvidia 的顶尖工程师还在。客户关系还在。但最主要的是,需求还在。

the demand for compute has not disappeared because this mother lover couldn't keep it in his pants. And tech didn't become obsolete because he had an affair. Kura is still there.

The networking effects are still there. The elite engineers in Nvidia are still there. The customer relationships are still there. But mainly the demand is still there.

无论你怎么看待 Jensen 这个人,他打造了一家非凡的公司,而非凡的公司生来就能超越其创始人长久存续。那么,我们该为波动做准备吗?绝对他妈该。我会具体说明为什么。但长期逻辑绝对完好无损。

Whatever you think about Jensen as a human being, I mean he built an extraordinary company and extraordinary companies are built to last beyond their founders. Now should we prep for volatility?

Abso freakingutely. I'm going to show exactly why. But the long-term thesis is absolutely intact.

11 月 4 日和 5 日,我们有三名共和党众议院委员会主席给 Jensen 发了一封信。外交事务委员会的 Brian Mast、中国委员会的 John Mullenar,以及情报委员会的 Rick Crawford。

他们希望他参加 11 月 18 日的闭门简报会。在信中,他们把一位名叫 Alice Wong 的女性——与 Jensen 没有亲属关系——称为公开报道过的中国情报资产,基本上就是间谍。

November 4th and 5th, we got three Republican House Committee chairs send Jensen a letter. Brian Mast from Foreign Affairs, John Mullenar from the China Committee, and Rick Crawford from the intelligence.

They want him in a closed dooror briefing on November 18th. And in that letter, they call a woman named Alice Wong, not related to Jensen, a publicly reported Chinese intelligence asset, basically a spy.

于是,11 月 8 日周日,Jensen 基本上给每一位 Nvidia 员工发了一封全员邮件,说:“听着,我对妻子不忠。”我的妻子。同一天,董事会宣布:“哦,我们要启动一项独立调查,由独立董事组成的委员会,并聘请外部律师。”

So Sunday, November 8th, Jensen basically sends out an all email to every single Nvidia employee says, "Look, I was unfaithful to my wife." My wife. Same day the board announces, "Oh, we're launching an independent investigation, a committee of independent directors with outside counsel."

周一开盘,该股收跌约 4%。当天晚些时候,甚至不到 24 小时,纽约时报报道称,商务部对一家名为 Mega Speed 的公司的调查在 Nvidia 介入后完全中止了。

Monday, we open the stock closes down about 4%. And later that day, not even 24 hours, the New York Time reports that the Commerce Department probe into a company called Mega Speed was absolutely halted after Nvidia intervened.

再次,纽约时报报道:“嘿,我们拿到了第二条消息。Jensen 据称与 Nvidia 医疗健康副总裁 Kimberly Powell 有染,她当时在 Jensen 手下工作。我是说,字面意义上,看起来就是直接向他汇报的人。”

Again, the New York Times reports, "Hey, we got a second story. Jensen allegedly had a relationship with Kimberly Powell, Nvidia's VP of healthcare, who was working underneath Jensen.

I mean, quite literally, it seems like someone who reported directly to him.

关于间谍那件事,你知道的,还有公司内部与下属的关系,当天晚上华尔街日报基本上报道说,看,董事会正准备在周四罢免他,他抢先一步,辞职了。我不认为他有多少选择余地,但他辞去了 CEO 职位,显然董事会即刻生效,Klet Crest 的 CFO 成为临时 CEO,基本上 Harvey Jones 成为 Nvidia 历史上第一位董事会主席。

There's one thing you know about the spy story like and inside the company relationship with his subordinate and that evening the Wall Street Journal basically reports look the board is prepping to oust him Thursday he gets ahead of it he resigns I don't think he was left a lot of choice but he quits the CEO position obviously the board basically effective immediately the CFO of Klet Crest becomes the interimm CEO basically Harvey Jones becomes the first chair of the board in Nvidia's history

周四晚些时候,路透社基本上说,看,我们又拿到了第三段关系,涉及另一名员工,而且 Jensen 已经同意在国会作证。所以,从董事会完全有信心、认为没有不当行为,到这个人成为历史,只用了 4 天。他走了。

and later Thursday Reuters basically says look we got a third relationship with another employee and Jensen already agreed to testify before Congress. So we went 4 days from the board has full confidence that no wrongdoing to this man is history. He's gone.

Alice Wang,又名 Wang Lei,与 Jason 没有亲属关系,只是同姓,但这基本上是个巧合。所以,她在 2023 年大约有八个月时间担任一家名为 Megape 的公司的董事总经理。

Mega Speed 是一家总部位于新加坡的公司,在东南亚建设数据中心。这些数据中心服务于中国客户,比如 Alibaba 这样的公司。

Alice Wang, also known as Wang Lei, no relation to Jason, same last name, but that's pretty much a coincidence. So, she was the managing director of a company called Megape for about eight months in 2023.

Mega Speed is a Singapore-based company and it builds data centers in Southeast Asia. And these data centers serve Chinese clients, companies like Alibaba.

向中国公司出租算力实际上是合法的。这是允许的。问题是,如果那些被禁的 Nvidia 芯片中有一些被非法运入中国本土,绕过并规避出口规则,那该怎么办。

Renting computer power to Chinese companies is actually legal. That's allowed. The question is what if some of those Nvidia chips that were banned were illegally shipped into China itself getting around and circumventing the expert rules.

根据《纽约时报》,基本上,Alis Wang 担任 CEO 的那家公司 Mega Speed 在两年内购买了约48亿美元的 Nvidia 芯片。这芯片数量可真不少。

Now according to the New York Times basically Mega Speed the company that Alis Wang was the CEO of bought about $4.8 billion of Nvidia chips over two years. That's a lot of chips.

Nvidia 的立场是,此前针对 Mega Speed 的转移指控已经过调查并被推翻。一切都没问题。中国方面说:“那些间谍指控全都是毫无根据的。”而这位女士 Alice,她目前在任何国家都还没有被起诉。

And Nvidia's position is that earlier diversion allegations against Mega Speed were investigated and disproven. Everything is fine. China says, "Oh, the whole spy claims are groundless."

And this woman, Alice, she hasn't been charged with anything yet in any country.

那么,我们从这个故事里究竟知道什么?我们知道美国情报部门怀疑她。我们知道 Jensen 和她有过一段关系。差不多就这些。她大概率是间谍吗?是的。这大概率是一个中国美人计吗?是的。

他是不是像个傻瓜一样掉进去了?是的。好吧。但再说一次,目前还没有100%的确认说这事已经板上钉钉。

So, what do we actually know from this story? We we know US intelligence suspects her. We know Jensen had a relationship with her. That's pretty much it. Is she most likely a spy?

Yes. Is this most likely a Chinese honeypot? Yes. Did he fall into it like a bone head? Yes. Okay. But again, there's no 100% confirmation that it's pretty much a done deal right now.

第一条,个人行为。我们看到了婚外情。我们看到了与直接下属的关系,这非常糟糕。正是这件事终结了 Jensen 的职业生涯。这一点毫无疑问。说实话,也应该如此。董事会尽到了自己的职责。

我的意思是,你不能和你的下属搞关系。你不能经营那么大一家公司,还树立这么糟糕的榜样,对吧?但业务本身,这只是噪音。眼下客户买 GPU,并不是因为 Jensen 的婚姻状况如何。

Story number one, personal conduct. So, we got affairs. We got a relationship with a direct report, which is horrendous. That's what ended Jensen's career. Make no mistake about it.

And honestly, it should. The board did its job. I mean, you can't have relationship with your subordinate. You can't run a company that size and be such a horrible example, right?

But the business itself, it's noise. Customers don't buy GPUs because of Jensen's marital status or lack thereof at this point.

第二条,国家安全呢?这个人有没有被拉下水?他有没有泄露任何机密?那些枕边风有没有导致什么本不该外泄的东西流出去?董事会的说法是,不不不。我们没有理由相信国家安全或 Nvidia 的机密曾被泄露。但再说一次,我们并不知道。我们对此还一无所知。

Story number two, what about national security? Was this dude compromised? Did he leak out any secrets? Did that pillow talk led to anything? basically that wasn't supposed to get out.

Well, the board is saying no, no, no. We have no reason to believe that the national security or Nvidia secrets were ever compromised. But again, we don't know. We don't know anything about that yet.

然后是第三条,Nvidia 的出口合规是否受到了损害。基本上,现在 Mega Speed 案已经作为刑事调查移交给了司法部。而如果最终查明 Nvidia 芯片被非法转运进入中国,那可能意味着罚款,可能意味着更严格的出口规定,也可能意味着短期内收入减少。

在这整场混乱中,这差不多是唯一可能影响公司的事情。

And then you got story number three, whether the Nvidia export compliance was compromised. Essentially, now the Mega Speed case has gone to the Justice Department as a criminal inquiry right now.

And if it turns out Nvidia chips were routed into China illegally, that can mean fines, that can mean tighter rules on exports, and that can mean in the short term less revenue.

That's pretty much the only thing here that may impact the company out of this whole chaos.

当我遇到股票的坏消息时,比如现在 Nvidia 这种情况,我喜欢把它分成三类:是噪音、是暂时的痛苦,还是基本上会推翻我对这家公司整个投资逻辑的重大严重问题。婚外情,这差不多就是噪音。

那只是他自己的过错和可耻行为,但这是私事。CEO 离职,是的,那是痛苦,但是暂时的痛苦。出口调查可能是这里唯一严重的问题,也是你必须密切关注的那个。但再说一次,目前什么都还没有。

Now, when I get bad stock news, like right now with Nvidia, I like to sort it out to three buckets. whether it's noise, temporary pain, or a major serious issue that basically undermines my entire thesis about the company.

Now, the affair, that's pretty much noise. That's just his own wrongdoing and shameful behavior, but it's a personal thing. The CEO leaving, yeah, that's pain, but that's temporary pain.

The export investigation is potentially the only serious issue here and that's the one you have to keep an eye on. But again, nothing yet.

周四,Nvidia 一度下跌16%,随后回升,收盘下跌4.3%。想一下这件事。公司创始人因间谍丑闻辞职。我不想说是什么性质的间谍丑闻,但他肯定是在搞间谍那套,而股价收盘下跌了4%。

Thursday, Nvidia went down 16% at some point, then recovered, closed down 4.3%. Think about it for a second here. the founder of the company resigned in a spy scandal. I don't want to say what kind of a spy scandal, but he was definitely pumping the spy and the stock closed down 4%.

整个一周,大约5000亿美元的市值被抹去。并不是说资金离开了 AI 板块,大家都收拾玩具回家了。就像我说的,看看股市正在发生什么。当 Nvidia 经历它那可怕的“高光时刻”——算是15分钟的公开羞辱——AMD 上涨了6%,Broadcom 上涨了6%。

所以,资金并没有离开 AI 板块。也没有离开半导体。它只是转向了目前没有丑闻的公司。说实话,这不是一个坏信号。如果投资者是在逃离 AI 革命,那就是完全另一回事了,但情况并非如此。

Over the whole week, about 500 billion in market value got erased. it's not like the money has left the AI sector and everybody just went home with their toys. Like I said, look at what's going on with the stock market.

As Nvidia was having its horrendous moment in the sun, it's kind of 15minute walk of shame. AMD was up 6%, Broadcom was up 6%. So, the money didn't leave the AI sector. It didn't leave semiconductors.

It just switched over to companies with no scandals right now. That's not a bad sign to be honest. If investors were running away from the AI revolution, I mean that would be a whole different story, but that's not the case.

现在,很多人不知为何认为这是 AI 泡沫破裂的开始。我的意思是,如果真是那样,你基本上会看到所有东西一起流血。你会看到 Nvidia、AMD,所有公司都疯狂抛售。我们并没有看到这种情况。

好吧。我认为需求逻辑完全没有被触动。市场基本上只是在说:“好吧,我们不知道这个板块里谁在经营最好的公司,而且因为有这么多不确定性,我们要退后一步观望。”市场绝对讨厌不确定性,这没关系。这完全没关系。

Now, a lot of people think for some reason that this is the start of the AI bust. I mean, if that was the case, you pretty much see everything bleeding together. You'd see Nvidia, AMD, all of them selling off like crazy.

We didn't see that. Okay. I think the demand story is absolutely untouched. The market is just basically saying, "Okay, we don't know who's running the best company in the space, and because we have all this uncertainty, we're going to take a step back."

The markets absolutely hate uncertainty, and that's okay. That is definitely okay.

这里有一个教训,远比 Jensen、Ellis Wang 或整个 Nvidia 要大得多。没有人。我向你保证,没有人的模型里预见到了这件事。没有人。不是高盛。更绝对不是摩根士丹利。

也不是 X 上那个眼睛冒激光的人。那还存在吗?人们还在往眼睛上加激光吗?我不知道。你可以做遍全世界的所有研究。你可以真的读遍每一份10-K。谢谢。你可以去参加每一场财报电话会,而有时候事情就是会从意想不到的地方冒出来,把你和整个市场都搞得晕头转向,对吧?

事情就是会发生,有时候根本没法预测。没有人的宾果卡上写着 Jensen 会有三段婚外情,其中一段还是和一个间谍。

there's a lesson to be learned here that's way bigger than Jensen or Ellis Wang or Nvidia in general. Nobody. Nobody, I promise you, had this in their model. Nobody. Not Goldman.

Definitely not Morgan Stanley. Not the guy on X with laser eyes. Does that still exist? Do people still put laser eyes on next? I don't know. You can do all the research in the world.

You can literally read every single 10K. Thank you. And, you know, go and attend every single earnings call and sometimes this is going to come out of left field and just bamboozle the hell out of you and the entire market, right?

stuff happens and sometimes there's literally no way to predict this. Nobody had this on their bingo card that Jensen is going to have three extra marital relationships, one with a spy.

如果 Nvidia 占了你全部净资产的60%,过去几天你肯定睡不好觉。问题不在 Jensen。当然,他做了这些事是个傻瓜,道德上应受谴责,绝对是个混蛋行为,但他不是这里的问题。

问题是你。问题在于你的仓位规模,因为这是你的投资组合,不是 Jensen 的。

If Nvidia was 60% of your entire net worth, you wouldn't had a good sleep over the past few days. The problem isn't Jensen. Of course, he's a bonehead for doing what he's done and morally reprehensible and absolutely a douchebag move, but he's not the problem here.

The problem is you. The problem is your position sizing because it's your portfolio. It's not Jensen's.

这不在我卖出公司的有效理由清单上,单凭这一点并不构成理由。这有很大区别。好吧。关键人物风险是真实存在的,但我也说过多次,它并不是自动全部清仓的信号。你必须评估这会对公司造成多大影响。

很多时候,联合创始人离开是重新评估你投资逻辑的好时机,确保一切仍然成立,但这绝对不是恐慌的触发点。它不是自动卖出的触发点。

It's not on my list of valid reason to sell a company just on its own. There's a huge difference. Okay. Keyman risk is real, but I also said multiple times, it's not an automatically full exit.

You have to evaluate how much that is going to impact the company. A lot of the times the co leaving is a good time to re-evaluate your thesis and make sure everything is in place, but it's certainly not a trigger to panic. It's not an automatic trigger to sell.

顺便说一句,先把话说清楚,Nvidia 不是 Palunteer。这里的 CEO 很关键,但与 Palunteer 不同,这里的 CEO 并不是唯一的东西。提醒一下,Nvidia 不是 Palanteer。

这是一家市值4万亿美元的公司。它有大量订单。它有大量积压订单。好吧,那些订单已经在那里了。它们不会因为 CEO 的名字变了就消失。

And by the way, just to kind of put it out there, Nvidia isn't Palunteer. The CEO here is key, but unlike Palunteer, the CEO here is not the only thing. Nvidia, mind you, is not Palanteer.

This is a $4 trillion company. It has massive orders. It has massive backlog. Okay, those orders are already in place. They're not going to disappear because the name of the CEO changed.

顺便说一句,看看 Nvidia 的董事会是怎么处理的。同一天任命临时 CEO,同一天任命新董事长。寻找正式 CEO 的工作,抱歉,已经在进行了。他们没有留下任何真空。他们是有继任计划的。

所以,是的,是的,是的。Nvidia 存在关键人物风险。但它没有人们想的那么重要。当然也没有 Palanteer 那么重要。

By the way, look at how the board handled it in Nvidia's case. Inter co name same day, new chair, same day. A search for a permanent C CEO, sorry, is already happening. They haven't left any vacuum.

You know, they got a succession plan. So, yes, yes, yes. Keyman risk exists at Nvidia. It doesn't matter as much as people think. Certainly not as much as Palanteer.

在左栏,列出你投资 Palunteer 的理由。列出你投资 Nvidia 的理由。在右栏,每个理由今天处于什么状态?好吧,我当初投资是因为 Jensen 是 CEO 吗?Jensen 走了。

好吧,那可能会改变一些事情。改变了多少?有多少没有变?

On the left column, list the reasons you've invested in Palunteer. List the reasons you invested in Nvidia. On the right column, what's the state of each reason today? Okay, so did I invest because Jensen is the CEO?

Jensen is gone. Okay, that may change things. How much of it changed? How much is is unchanged?

第三步,你得像个购物者一样思考。现在,Nvidia 的市盈率是多少,14、15倍。这太疯狂了。如果你知道,如果苹果给你一部 iPhone,全新的 iPhone,打40%、30%的折扣,你会买三部。为什么换成一家伟大的企业就不一样了?想一想。

Step three, you got to think like a shopper. Right now, Nvidia is trading at what 14 15 PE. It's insane. If you know, if if Apple gave you an iPhone, brand new iPhone at a 40, 30% discount, you'd buy three of them.

Why is it different with a great business? Think about it.

左栏。你当初为什么要持有 Nvidia?好吧,显然我投资逻辑里的第一个理由是 AI 需求。世界还在建设 AI 吗?是的,绝对没有改变。第二,地球上最大的那些公司正在数据中心上花费数千亿美元。

仍然成立,没有改变。好吧,到目前为止一切顺利。CUDA 呢?软件生态系统。地球上每一个 AI 开发者都是在这个生态上训练的,没有改变。第四,网络。把数千块 GPU 连接成一台巨型计算机的东西,没有改变。

第五,人才。数千名杰出的工程师,没有改变。产品路线图,没有改变。那么他们改变了什么?好吧。第七,他们变了,创始人 CEO 要离开了。显然,这是一个变化。

Left column. Why would you own Nvidia in the first place? Okay, obviously the number one reason in my thesis is AI demand. Is the world still building AI? Yes, absolutely unchanged.

Number two, the biggest companies on Earth are spending hundreds of billions of dollars on data center. Still true, unchanged. Okay, so far so good. What about CUDA? The software ecosystem.

Every single AI developer on the planet is trained on unchanged. Number four, networking. The stuff that connects thousands of GPUs into one giant computer, unchanged. Number five, talent.

Thousands of brilliant engineers unchanged. The product road map unchanged. So what did they change? Okay. Number seven, they change a founder CEO is leaving. Obviously, that's a change.

那因为潜在的出口违规而带来的监管风险呢?是的,这一点我们得盯着。我们目前还不清楚,但没错,这确实可能是个问题。

What about some regulatory risk because of the potential export violations? Yeah, we got to watch that. We still don't know, but yeah, that certainly be it.

所以,有六个和上周完全一样,另外两个里,基本上一个是需要关注的项目,另一个已经发生了变化。

So, you got six that are exactly where they were last week and two that basically one is a watch item and one has changed.

在我看来,恕我直言,就现在来看,我可以坐在这里坦率地告诉你,这是一个被打破的投资逻辑。好吗?它上面确实有一道疤。这里有个瑕疵。有一道划痕。再说一次,这就是为什么你能拿到折价。风险确实更大。但这是不是一个被打破的投资逻辑?我不这么认为。

To me, with all due respect, looking at this right now, I can sit here and honestly tell you this is a broken thesis. Okay? It definitely has a scar on it. There's a there's a blemish here.

There's a scratch. And again, that's why you're getting the discount. Certainly has more risk. But is this a broken thesis? I don't think so.

现在,我想在这里更深入地谈谈需求,因为需求在这里就是全部的关键。这已经成了定局。它正在发生。所以,你觉得明天 Google 的 CFO 会走进会议室说:“各位,我们得取消这个,因为 Jensen 出了轨”?这不会发生。

Now, I want to go deeper here on demand because demand is the whole game here. It's already done. It's happening. So, do you think tomorrow the CFO of Google is going to walk into a meeting and say, "Well, guys, we need to cancel this because Jensen had an affair." It's not going to happen.

算力正在变得更便宜。技术会变得更便宜。人们用得更多,需求就会上升。这就是杰文斯悖论。蒸汽机,你去 Google 一下。更便宜的 AI 意味着更多的 AI 需求,意味着更多的芯片。

对算力的需求不会因为 Jensen 的个人选择而消失。它会上升。总得有人来填补这个空缺。而 Nvidia 无论有没有 Jensen,都处在完美的位置上。

Compute is getting cheaper. Technology gets cheaper. People use it more, the demand goes up. That's literally the Jevans paradox. The steam engine, Google it. Cheaper AI means more AI demand means more chips.

The demand for compute is not going away because of Jensen's personal choices. It's going to go up. Somebody has to fill this void. And Nvidia is in the perfect position with or without Jensen.

现在说说 Nvidia 的护城河。如果你以为 Jensen 那件酷酷的皮夹克就是护城河,抱歉打破你的幻想,并不是。他穿着那件皮夹克、带着他那些花招在台上确实是个很酷的人。但人们买 Nvidia 的 GPU 并不是因为他那件该死的皮夹克。

好吗,护城河其实就是 Nvidia 的生态系统。

Now, Nvidia's moat. Now, if you think that Jensen's little cool leather jacket was the moat, sorry to burst your bubble, it wasn't. He was a cool dude to watch on stage with that leather jacket and with him with his shenanigans.

But people were not buying Nvidia GPUs because of his freaking leather jacket. Okay, the mode is literally the ecosystem of Nvidia.

把 CUDA 想象成铁轨的宽度。每一个 AI 开发者、每一个实验室、每一所大学、每一家初创公司,都必须把自己的火车造得能适配 Nvidia 的铁轨。

Think of CUDA like the width of railroad tracks. Every single AI developer, every lab, every university, every startup, they have to build their trains to fit the Nvidia tracks.

你能现实地切换到竞争对手那里吗?当然。我是说,这是可能发生的。理论上有可能。你只需要把每一列该死的火车都重造一遍。把每一条该死的铁轨都重新铺一遍。重新培训每一位火车工程师,然后只能祈祷别出问题,因为一切都是全新的。

从 Nvidia 切换出去的成本高得远超海格力斯级别。说海格力斯级别,大概是今年最轻描淡写的说法了。这就像赤身裸体、单手绑在背后去爬珠穆朗玛峰。画面可不好看。

Can you realistically switch to a competitor? Sure. I mean, it can happen. It's possible theoretically. You just have to rebuild every freaking train. Relay every freaking track. retrain every single train engineer and they just hope nothing breaks because it's brand new.

The switching costs from Nvidia are beyond herculan. Herculan is probably the understatement of the year. It's like climbing M rivers naked with one hand and tied behind your back. Not a pretty sight.

各位,Jensen 卸任之后,这一点并没有改变。然后还有网络效应。想想看。我是说,把成千上万块 GPU 连接起来、让它们在一个数据中心里协同运作的能力。拜托。让这一切在现实中发生的,可不是穿着皮夹克的 Jensen。

That didn't change when Jensen stepped down, guys. Then you've got the networking effects. Think about it. I mean, the ability to connect these thousands of GPUs and make them act all in one data center.

Come on. It's not Jensen who's making that happen physically out there with his leather jacket.

现在,很多人问我:“那 Nvidia 没有 Jensen 能活下来吗?”我们来谈谈这个。我们对自己诚实一点。

Now, a lot of people are asking me, "Well, can Nvidia survive without Jensen Tom?" Let's talk about it. Let's be honest with ourselves.

我是说,这个教训其实并不是“创始人无关紧要”。我想这个教训更多是:一家伟大的企业会存活下来,并且比它的创始人活得更久。它本来就是为此而设计的。

I mean the lesson really isn't oh founders don't matter. I guess the lesson is more a great business is going to survive and outlive its founder. It's designed to.

好,这就是为什么现在你真正最需要关注的是联合搜索,而不是去读那些关于他跟谁上床的小报头条。恕我直言,Nvidia 是一家非常不寻常的公司。有超过 60 位高管直接向 Jensen 汇报。

对一家市值 4 万亿美元的公司来说,这至少可以说是很不寻常。呃,几乎没有人这样经营公司。好吗。Jensen 真的真的深入到每一个项目、每一张路线图、每一个客户的细节里。他他深入到所有事情的细节里。

Okay, that's why the co search is probably the most important thing you have to watch right now instead of reading all this tabloid headlines about who he was banging or not. With all due respect, now Nvidia is a very unusual company.

More than 60 executives reported directly to Jensen. That is unusual to say the least for a $4 trillion company. Uh almost nobody runs a company like that. Okay. Jensen was really really in the weeds of every single project, every single road map, every single customer. He he was in the weeds of everything.

所以百分之百,我可以告诉你,下一任 CEO 不会照搬 Jensen。他们得重建组织结构,增加更多层级,增加更多,你知道,更多这种,我猜是这类公司更传统的结构,但无论有没有 Jensen,这都会发生。

就算他十年后辞职、直接退休,这也会发生。Nvidia 需要做到这一点。

So 100% I I can tell you that the next co isn't going to copy Jensen. They're going to have to rebuild the structure, build more layers, more you know more this I guess traditional structure of this sort of company, but it's going to happen with or without Jensen.

It would have happened if it quit in 10 years and just retired. That needs to happen with Nvidia.

Jensen 是独一无二的,而现在他走了。接替他的那位女士将担任临时 CEO。直到昨天,她还是 CFO,也就是首席财务官。她在公司待了大概 14 年。所以,她对这家公司财务的每一个细节都了如指掌。

她认识每一个客户。所以,在一段时间内,她会是一个稳健的掌舵人。

Jensen was one of a kind and now he's gone. Now, the woman who replaces him, she's going to be the interimm CEO. Until yesterday, she was the CFO, the chief financial officer. She's been with the company for like 14 years.

So, she knows every single iota in this company's financials. She knows every customer. So, she's going to be a steady hand for a while.

她会成为新任 CEO 吗?大概率不会。大概率不会。他们会去寻找一个能驾驭这一切的人。但我喜欢我目前从这个过程中看到的东西。

Is she going to be the new CEO? Probably not. Probably not. They're going to be searching for somebody that can handle this. But I like what I see from the process so far.

好。你同不同意董事会的做法,那是另一个问题。好。你可能会说:“董事会让 Jensen 走简直是白痴。”哦,不,董事会,他们本来就该这么做,因为这件事在道德上实在太不可接受了。无论你怎么看他们所做选择的质量,有一点你必须同意,那就是他们行动得该死的快。

好,他走了。同一天你就有了临时 CEO,同一天他就退出了董事会,同一天你就有了新的董事长,同一天新的搜寻就启动了。我是说,他们没有僵住。这里有一个清晰的计划,有一个清晰的继任流程正在推进。

Okay. Whether you agree or disagree with the board, that's a separate question. Okay. You may say, "Well, the boards are idiots for letting Jensen go." Oh, no, the boards the board, they should have done it because it was so immorally irrehensible.

Whatever you think about the quality of the choice they've made, one thing you have to agree on is they move freaking fast. Okay, he gone. Same day you got an interim CEO, same day he's out of the board, same day you got a new chair, same day the new search starts.

I mean, they didn't freeze. There is a clear plan here and there's a clear succession process that's happening.

现在,如果你看看当前的估值,我们在这里谈点数字。呃,那个 4 倍市盈率,基本上就是你今天所支付的价格相对于公司明年预期盈利的比值,低于 14。呃,只是为了让你感受一下这有多疯狂。

好,这让 Nvidia 的远期市盈率比 UPS 还便宜。一家包裹递送公司现在比 Nvidia 还贵。地球上最重要的 AI 公司,定价居然像一家送快递的公司。

Now, if you look at the valuation right now, and let's talk numbers here a little bit. uh the 4 PE which is basically what you're paying today compared to what the company's expected to earn next year is below 14.

Uh just to give you context of how insane this is. Okay, that puts Nvidia cheaper on forward PE than UPS. A package delivery company is now more expensive than Nvidia. The most important AI company on Earth is literally priced like a company that delivers boxes.

与此同时,过去12个月的营收约为3000亿,增长了约83%。净利润约为1930亿。这些数字极其出色。我说这些数字出色,都算是轻描淡写了。

Meanwhile, revenue over the last 12 months is about 300 billion up around 83%. Net income is about 193 billion. Those are beyond excellent numbers. I saying those are excellent numbers is going to be an understatement.

而这只股票目前仍比他们10月份创下的历史高点低35%。

And still the stock is now 35% below the October record that they had.

市场总是会把未来的风险当作已经发生来定价。眼下,市场基本上是在按英伟达的领导地位已经彻底失败来定价。就好像客户全都走了,专家案基本上要爆掉,所有人都会进监狱,整个公司会被解散。

是啊,当然。如果这些事真的发生,那也行。但价格已经反映了其中很多因素。这些事大概率不会发生。

The market is always going to price future risk as if it already happened. Right now the market is basically pricing Nvidia like the leadership position basically failed. Like customers have all left and the expert case basically is going to blow up and everybody's going to prison and they're going to disband the entire company.

Yeah, sure. If those things happen, fine. But the price already reflects a lot of this. That's probably not going to happen.

我把这件事更多看作助燃液,而不是别的什么。为什么是助燃液?如果你曾经试着生火,显然,想让事情轻松一点,你就需要一些助燃液,对吧?没有助燃液也能行。你可以用它来加快速度。那如果没有木柴、只有大量助燃液呢?

你会搞出一场大爆炸。几英里外的人都能看到天上的火光。但这场火能烧多久?它会非常、非常快地烧完。在这个比喻里,木柴是什么?盈利、利润率,它们已经烧了好多年,而且其中没有任何东西受损。

所以,我们只是看着助燃液在燃烧。烟花放得惊天动地。谁在乎呢?

I call this thing more of a lighter fluid than anything else. Why lighter fluid? Now, if you ever tried to start a fire, obviously you're going to need some lighter fluid if you want your life to be easier, right?

You can do without lighter fluid. You can, you know, you can speed things up with it. What about zero firewood and a lot of lighter fluid? You're gonna have a big kaboom. People miles away going to see that thing in the sky.

But how long is this fire going to last? It's going to burn out very, very quickly. What is the firewood in this analogy? Earnings, margins, they burned for years and nothing about them has been damaged.

So, we're just watching lighter fluid burn. The fireworks are happening crazy. Who cares?

当然,有一些风险值得关注。好吗?我不是视而不见。承认某些风险是没问题的。11月17日我们会迎来财报,我预计会非常、非常好。但最大的问题是,他们会下调指引吗?这会很有意思。

Now, there are certain risks worth watching. Okay? I'm not blind. It's okay to acknowledge certain risks. On November 17th, we're going to have earnings, which I expect to be very, very well.

But the big question, do they cut guidance? Interesting conversation.

那整个 mega speed 和司法部的故事呢?哦,那可能是个大事件。如果专家规则变得更严格,或者英伟达受到处罚,那就看看这事会怎么发展吧。

What about the whole mega speed and justice department story? Oh, that could be a big one. You know, if expert rules get tighter or Nvidia gets hits with penalties, let's see how this plays out.

那董事会自己的调查呢?谁在什么时候知道?好吧,很多问题,很多需要解释的地方。

What about the board's own investigation? Who knew when? Okay, a lot of questions, a lot of explaining.

当然,还有国会。他们会要求 Jensen 作证。那 CEO 人选呢?他们会选个笨蛋吗?大概率不会,但这也是另一个风险。

Of course, you got the Congress. They're going to have Jensen testify. What about the CEO pick? Will they pick a bozo? Probably not, but another risk.

现在,我跟你说实话。在 Jensen 辞职之前,很多华尔街分析师基本上都说:“看,这可能会让股价下跌20%。”而且,未来几周我们肯定可能看到更多抛售压力,在调查、交接、整个媒体闹剧之前。

我不认为这只股票未来几周会走直线。它会波动。你应该为波动做好准备。

Now, I'm going to be honest with you. Before Jensen resigned now, a lot of Wall Street analysts basically said, "Look, this could take the stock down 20%." you know, and certainly we could see more selling pressure in the coming weeks, you know, before the investigation, the transition, the whole media circus.

You know, I don't think we're going to have a straight line with this stock in the next few weeks. It's going to be volatile. You should prep for volatility,

但唯一会让我改变对这只股票看法的事情,是营收增长开始逐季放缓。如果利润率开始呈趋势性收缩,如果客户在流失,那才是问题;而我们现在看到的所有这些其他事情,都只是噪音。

but the only thing that would make me change my mind about this stock is if revenue growth starts slowing down quarter after quarter. If margins start shrinking as a trend, if customers are leaving, not this, all of this other stuff that we're seeing right now is just noise.

所以,我们要务实,而不是情绪化。如果你已经持有英伟达,而且仓位规模合理,你就应该继续持有。你不应该因为新闻头条就卖出。不要因为价格下跌就卖出。也不要因为价格下跌就买入。在这里卖出意味着你必须对两次:卖出的时机和重新买入的时机。

So, let's be practical instead of emotional. If you already own Nvidia and your position size makes sense, you should hold. You shouldn't sell on headlines. Don't sell because the price dropped.

Don't buy because the price dropped. You know, selling here means you got to be right twice. Right on the exit and the re-entry point.

如果你想买入英伟达,现在就是你的机会。一个更好的价格。你应该做美元成本平均法,也就是定投。不要一次性全投进去。你不知道这东西的底部在哪里。不管它现在看起来多便宜。一次性投入就是一场赌博。

不要赌。把投入分散到未来12个月。你现在有一个不错的入场点,用定投稍微利用一下。

If you want to buy Nvidia, here's your chance. A better price. You should do dollar cost average DCA. Don't lump sum in. You don't know where the bottom of this thing is. Don't don't care how cheap it seems right now.

Lumpsuming is a bet. Don't take a bet. Spread it out over the next 12 months. You got a good entry point right now. use it a little bit with DCA.

现在,按照我的系统,英伟达已经远低于52周高点的20%阈值。所以,在我的系统里,这属于加倍买入的区间。在我的学院、我的投资系统里,大家都知道这一点。各位,你必须坚持自己的计划。

Now, for my system, Nvidia is way more below that 20% threshold of the 52- week high. So, in my system, that's double down territory. Everybody knows this in my academy, in my investing system. And you got to stick to your plan, guys.

如果你现在惊慌失措,对自己关于这只股票的投资论点失去信心,那说明你从来就没有过投资论点。看,Jensen 的个人选择丝毫不会削弱正在我们眼前发生的 AI 革命。我对 AI 革命的信念没有改变。

我对英伟达竞争优势的信念没有改变。对英伟达的长期看涨论点没有改变。

If you are freaking out and losing your confidence in your thesis about this, that means you never had a thesis. Look, Jensen's personal choices do not take one iota away from the air revolution happening right in front of us.

You know, my conviction is in the air revolution not change. My conviction in Nvidia's competitive advantages did not change. Long-term bullish thesis on Nvidia did not change.

市场能够、也一定会因为不确定性和这类行为而惩罚股票,而有时这会给有耐心的投资者创造出绝佳的买入机会。这就是为什么沃伦·巴菲特总说,股市是一个把财富从不耐烦者转移给耐心者的机制。

如果英伟达的基本面保持住,我相信这可能是这轮 AI 周期中最好的买入机会之一。

The market is able and will punish stocks for uncertainty and this sort of behavior and sometimes that will create incredible buying opportunities for patient investors. That's why Warren Buffett always said the stock market is a transfer mechanism from the impatient to the patient.

If Nvidia's fundamentals hold, I believe this could be one of the great buying opportunities of this AI cycle.

观察点

NVDA 11月17日的财报与业绩指引
业绩指引被下调
11月17日NVDA 财报
业绩指引下调将削弱该论点
NVDA 未来几个季度的季度营收增长、利润率与客户需求
营收增长逐季放缓、利润率呈趋势性收缩,或客户流失
未来几个季度NVDA 季度财报
会让他改变想法,并削弱看涨论点
司法部对 Mega Speed 的调查以及 Nvidia 出口合规的结果
出口规则收紧或 Nvidia 受到处罚
未知司法部调查/出口合规调查结果
将削弱该论点

这个频道对 $NVDA 的历次判断

Tom Nash 在这只股票上共 3 条判断,这里只列相邻的几条。

跨洋听风